New Group Calls For Boycotting Systemd

[Michael_S]

I forget who shared this video, but it was quite good: https://www.youtube.com/watch?v=-97qqUHwzGM The first few minutes are generalities and comedy, if you want to skip to the meaty points of the argument it's 4 minutes in: https://www.youtube.com/watch?featur...qqUHwzGM#t=257

No. You do want to be able to repair a machine from remote and without taking the entire OS down. I suggest you first start working in system administration of a larger company and get some hands-on experience on how it is done.

By the way, not knowing why it fails is only bad. When you do not know then how can you stop it from happening again?! You cannot just leave it to luck.

I know your way of thinking is kind of common among Windows users, but when you look at Microsoft then you will see that even they want an OS, which can repair itself and does not require external tools every time it fails.

What you are talking about is not best practise. It is only the worst case scenario and you sure do not want to work in an environment where every incident is also the worst case and your solution is to take the hammer each time.

I disagree. If you know what took the system down, then repairing it in place is the way to go. But if you're not certain, your only safe option is a clean reinstall - hence the popularity of bare-metal-provisioning tools, so that "nuke from orbit and start over" is a few keystrokes and then you can work on something else.

[WizardGed]

Should have been paying closer attention to that wikipedia article.

FreeBSD doesn't uses SysV init, and they have no concept of run levels. All of the BSDs use BSD Init which is miles ahead of SysV, which is interestingly similar in style from the user facing side to systemd init.

You're absolutley right of course, I should have put solaris or AIX in BSD's place. I named BSD because it was most likely more familiar to him than the those other unixes. Sorry about that should have payed closer attention to my own link.

[sdack]

I disagree. If you know what took the system down, then repairing it in place is the way to go. But if you're not certain, your only safe option is a clean reinstall - hence the popularity of bare-metal-provisioning tools, so that "nuke from orbit and start over" is a few keystrokes and then you can work on something else.

Sure, but this is not what we were talking about. You are at best advocating a solution for the worst case. You do not want it to be the solution for the common case. You do want a system where you have a good chance of finding the cause of the problem and fix just that, because when you do not it and you always only "nuke from orbit" then you will find yourself out of options the moment your "nuke" itself is broken.

Your "nuke from orbit" is also not in any way new, because it is just a classic restore from a backup copy. So your "nuke" is technically already outdated, because modern backup solutions offer a wide range of options and more than just your "nuke". You can run checks, take snapshots of active partitions, restore onto active partitions, do incremental and selective backups, and all with GUIs and from remote.

So try not to look for the hammer. Better look for smaller tools and less intrusive solutions. And find the cause and do not just fix the symptom.

[blackiwid]

No. You are twisting things around, because you do not understand why we have them, and because spinning them around does not give you an answer is your conclusion to throw it all away. It is however not a sane approach to solving problems.

so u want to use gnome for the next 50 years and the 1-2 first shedulers also not the newer ones, and btrfs is also evil? because its not just another implementation of the same featureset but completly different, it makes the good old lvm obsolete and swraid and so on... and btw upstart is also evil. what else git is evil, because we had svn we could have fixed it, we even did that a big now. wayland is evil, we could have slowly fixed xserver, in 100 years I am shure it would be really good state.

We have always relied on human readability

in config files yes there its pretty important, except your software is so good that u never have to change the config settings because u have a very good config dialog. that depends on the use case for a gnome desktop I dont need to force the users to edit config files, for apache its good that there is config files in txt form, there is a big advantage to have them this way, because u can open it with a editor, but even there, as example cron creates only a temporary config file u edit, if that would be saved as binary I would not care to much but ok, for config files I see the point.

for log files, no there is no good reason to do that, the output is for humans not readable, except u have 1000hours free time, only the grepped files or otherly created output is usable.

as well as backwards compatibility and for good reasons. Is cat and grep a program? Yes.

we did care about backwards compatibility? No in LInux we always break everyting as much as possible, we get hate-flamewars on daily basis for breaking kernel ABIs, we often break proprietary software by new releases of xserver. every postgres or mysql upgrade breaks the old data or must be upgraded, its linux major thing that we do not care we break everything if we have a sane reason for that.we dont keep backward compatibility as long as possible, because if we change the driver api of grafic drivers we can in an eyeblink also change the drivers in the kernel, and thankfully at least on that level nobody cares about this blobs from companies.

Is text just a sequence of bits and bytes? Yes. It is however no reason to throw it away and to reinvent it differently, because then we can do the same with everything else, past, present and future software, including any new implementations. Nothing will stop us from doing it over and over again.

do we through away text, we use text where it makes sense and where it does not we dont use text, is text the holy grale to do things always? again if thats true we need to replace every binary programm and replace it with some interpreted source files and interpret them. You gave no reason why keeping log in txt files is a good thing. I see log files not very usable for normal people correct me, but for most linux systems they only grow and grow and grow till the filesystem is full. its a pain in the ass, having several differnt ones, we have dmesg already which gives most of the time the interesting output. its a mess even as 20 year linux user but mostly as desktop but having installed several gentoos compiled own kernels, this log system made never much sense to me. U have several files and mechanisms for the same thing, log the system and nothing does the job right, some stuff just happen before logging is starting, if systemd gives finaly ONE log from point 0 to end I am happy as hell.

It will simply not lead anywhere when it only leads in circles.

what do u mean by that, I have no glas bubble.

This is why we preserve as much of the old as possible,

do we? where is this idea coming from? we could use a 3.0 kernel still and only update the drivers nothing else. we release every 2 seconds a new kernel etc we replace as much code as possbile, we even messure the lines of codes of linux and its changelogs and am proud if its very high.

The opposite is true we are sad that we are so slow in changing things, we are said that only the kernel has enough manpower to change stuff fast enough.

why we only fix what is broken and why we build on top of it. Only once you understand this can you understand how progress is made.

I dont know where u have your ideals did rms tell something like that, or Linus or who? I never heard such stuff, we build on top of stuff becuase its easier, and most of us are lazy not more not less.

lets take vim, at some times u have to cut out old stuff. I tried to fix my software for months but at some point I also made a -ng version, that only ports some stuff from the old one into the complete new basis because u can come with fixing the old ones only so far. U can put on a bycycle a motor and call it pedileg, but u cant drive with it on the high way (or do u americans have something similar like "autobahn") u have at some point and maybe thats even before the bycycle gets made put 2 other wheels on it and add some case around it, so taht u can drive faster without dying automaticly if u get in a accident, some people dont get that they drive motorcycles, and they live very dangerous. its ok if they choose to do so, but for most people its no good solution.

yes there must be a good solution to break old stuff, u dont do it for no good reason, I was sceptical because I did not see as user so big advantages of newer init systems, but just with lenerds talk now I see HUGE advantages of systemd that can be accomplished. And even now on small things systemd is way better than everything else ever was, like distro-independend init files.

systemd works, it does its job well, but it is also a step backward.s. Once any of its dependencies become outdated and need replacing, and it will happen, will you see how much of a problem it actually is. Perhaps you now think we should not change it in the future and then only fix what is broken and then to build on top of it, but you will not do it, because what you do not learn now you also cannot use in the future.

I see risks in using systemd, but can it be worse, today I have to more or less use gnomestuff in i3wm, to get a good experience power management volume control... wlan control, is there ways around yes. are they very painless no. So putting many stuff gnome did into a lower abstraction layer is a good thing? YES!

Is it then impossible to bypass that? No I can do systemctl suspend but I still can do echo 1 > /sys/power/state (was it that file?)

To spin it as far as to think one could win against Microsoft shows how deep you have fallen into the illusion, because Microsoft software never has been great to begin with. However, they do know how to influence politics, users, sales, companies, etc. to become one of the biggest de-facto monopolists the economy has ever seen.

thats true but they were never 1000 km behind they were most of the time good enough a small bit worse here and there, but nothing that problematic, why is android so popular, becvauase people start to really extremly hate windows and thats not linux fanboys. why is apple so succesful, why are consoles so successful. yes linux could in the past not make much with the hate people have for windows. because linux does not fix many of the most important problems. but if gnu/linux not wins, google will win. So its the question is Android the next desktop system for the masses or gnu/linux, it cant be apple except they would release it for every pc and not their highprice pcs.

To believe that if only one could write perfect software, which does not exist by the way, one could also beat Microsoft is naive.

did I write something about perfect Software? I wrote several points, systemd alone will not kill microsoft, btrfs alone also not and steam alone not too, if we would have xserver the next 20 years people would not use it for desktop. but wayland alone will also not be the solution. But all this together, there is not much left except windows fanboys, but I dont think that most windows users are fanboys.

And yes btw all this alone will also not help, steammachines is a important point, because we need people that are not willing to risk to install it by them self, but no matter how good it linux is, having to install it manualy will stop us anyway. But there are companies like system76, and with steammachines at some point companies will come to fill the bigger demand of linux pcs.

Its not one point, but its ONE additional nail in the coffin of windows.

We will more likely see the desktop PC die before this happens, because while Linus still dreams of conquering the desktop have others been working around it for a long time now. Just see where Linux can be found these days. Most of us do not really want a dominance. We only want Linux and do not care for economical or political power.

so where is the problem, we have distries and if steamos is the distro for masses how does it hurt xy-linux distro? I even suspect a fork of linux in the long run. Because the differneces from the opensource guys and the free software guys are just to big.

In reallity we even kind of have that today, its maybe more of a patchset but whats there the difference to a fork, that we dont give him a new name, some fsf aproved distros use such kernels.

And at one point we may see taht. if systemd is really that bad we see a fork of that, till now its great.

And I also dont eager to much to be the most used os, its just the problem that I have to have a windows pc or nonfree console if I want to play, and a 2% operation system will not keep relevant for gaming companies in the long run. So it has to have some kind of success if its only 10-20% thats maybe much enough but I just wondered if there will be a holding I just doubt that windows can stay relevant, it lost most of its users already to android, and if they do not fix any of their problems I just dont see them keep relevant. And again I do care at least that much that I dont want to buy laptops with windows lisenses included, I want perfect linux support so that a new laptops fan control goes not nuts under linux etc...

And is there a way to go around this? buy a tablet + a keyboard dock if you are happy with android yes, or if u invest 500 extra dollars and half of your akku times to get a i3 processor because baytrail is optimised for windows 32bit and dont runs with a normal grub yes. there are painful ways to build around that.

Will I stop using linux when in 20 years it still has only 2-5% market share no, will I be said if we have 20-80% market share no, too.

[sdack]

so u want to use gnome for the next 50 years ...

I have been using UNIX systems for the last 25 years. I certainly do not see a problem using it for another 50 years. Just look at what it brought us. There is little wrong with it.

So if you want to discuss this with me then please respond properly, because I am not interested in a conversation where someone fails to understand my comment and then only tries to defy every single sentence I write, because he forgot how to respond. Read my comment as a whole, try to understand it, and then respond. I am not interested in picking the bones out of your comment.

[interested]

No. You do want to be able to repair a machine from remote and without taking the entire OS down. I suggest you first start working in system administration of a larger company and get some hands-on experience on how it is done.

By the way, not knowing why it fails is only bad. When you do not know then how can you stop it from happening again?! You cannot just leave it to luck.

I know your way of thinking is kind of common among Windows users, but when you look at Microsoft then you will see that even they want an OS, which can repair itself and does not require external tools every time it fails.

What you are talking about is not best practise. It is only the worst case scenario and you sure do not want to work in an environment where every incident is also the worst case and your solution is to take the hammer each time.

Well, I worked in the IT industry when MicroSoft Xenix was still around (Few people know that MS once had a real Unix OS), so I have seen many different admin solutions over the years.

So I stand by my point; if there are valuable data on machine that won't boot properly, you don't try to repair it from the OS itself, you boot from a rescue media. So the idea that an old FS layout and partition layout are a good idea because they _may_ help repairing a live OS with disc problems, is really a bad idea.

Sure, remote administration is the norm for many things, like helping a user with a printer problem, installing new software, diagnosing why email doesn't work, deploying new workstations or servers etc. But at the moment it is quite hard to remotely boot from a remote disc and mount the local disc as RO. Interestingly enough, it is a problem that the systemd group have long term plans on fixing (stateless boot and all that).

I have worked with the Police regarding various IT crimes, and seen my share of flaky raid controllers, driver bugs, badly terminated SCSI chains, and an endless row of failing hard discs, so I fully understand the background why it is considered best practise in the industry, to only mount broken discs as RO if you care about the data.

/sbin is a relict from the old days of doing computing, same with having a separate /usr. The reason why Linux are changing the OS FS layout with regular intervals, is because the old layout doesn't fit the new way of doing things. Same with systemd replacing SysVinit really.

[blackiwid]

I have been using UNIX systems for the last 25 years. I certainly do not see a problem using it for another 50 years. Just look at what it brought us. There is little wrong with it.

So if you want to discuss this with me then please respond properly, because I am not interested in a conversation where someone fails to understand my comment and then only tries to defy every single sentence I write, because he forgot how to respond. Read my comment as a whole, try to understand it, and then respond. I am not interested in picking the bones out of your comment.

Funny that u answer to one typo of me and ignore the 1000 other arguments and u said basicly 0 arguments at all.

I meant gnome2 not gnome.

But again u live in your special world the opposite world or whatever, I try to bring other arguments than, ... we do things this way and that way , when "we" means I personaly alone like it that way I not we.

But forgive me to even having tried to argue against some random text that included no arguments at all.

But be not to said, even normal people think that cloth made by cotton is a good thing maybe mix it up with hemp, there is always room for fetisch and there will be a latex-linux-from-scratch distro for u to get your kick where u can watch your ascii bits flow around as much as u like it.

[kringel]

I think many people miss the basics here: Red Hat is not just one specific distribution of GNU/Linux. Both Sievers (on the kernel side, e.g. udev and kdbus) and Poettering (userspace side, systemd and -afaik- also kdbus) work at Red Hat. It is by no means evil in itself, that they make systemd the way that it is the best fit and has the perfect use purpose for Red Hat's Enterprise Linux and Red Hat's customers (military and embedded stuff etc.) That is ok and quite logical when you think about it.

However the problem is that many think systemd would be just a replacement for an init system. It's not. Sievers called it "basic userspace building block to make an OS from" e.g., Systemd spreads into areas of the networking stack, util-linux, logging, and more will follow. It standardizes the tools, the toolchain solutions of *every* distribution which uses it (which are pretty much all distribution, except Slackware and Gentoo and maybe a new version of LFS).

The problem is: When I was your age those tools WERE the distribution. The were the heart of every individual distribution. Yes, more or less the same in different ways got achieved. Systemd will replace this over time. And Lennart's "vision" about Brtfs, subvolumes, the even attack different package management systems.

Ask yourself: What pieces will remain of your favorite distro? The wallpaper package maybe? perhaps!

You give all funtionality into the hands of systemd and its core developers and Red Hat. And systemd will grow even more. (We need a parody with systemd being the evil in the LoTR-FOTR opening scence, -IMHO-) That said of course it can be convient, of course it might attract more software developers to Linux because of this standardization. But systemd already is a beast which is so big and shitty programmed that no one understands what it does anymore. If I were a millionaire I would build an open source company to audit the shit out of systemd, just to pick at Red Hat, sadly I'm not a millionaire. When you have all the functionality of the above packages now concentrated in systemd, pretty much standardized and the same in all major distributions, this will become unmaintainable, and an extremely huge vulnerability and security risk.

Systemd is UNNECESSARILY so big as it is and with that it violates the Four Freedoms of GNU. That is no requirement of the GPL or the open source term. But it is a huge sign with huge letters saying "F*ck you!" to the whole Linux and open source community. They know better, you are about to shut up.

Next problem: You think Red Hat is just the opposite of Microsoft? Like Red Hat would be the good guys? Seriously? Red Hat is all about creating an INDIRECT VENDOR-LOCKIN with all their recent stuff. (I just don't get why all major distros and even Debian jump on the 'bleeding edge bandwagon' with this hyped and dangerous crap, when they still have a working init system.) Red Hat is about locking freedom GNU freedom. Poettering talks about "GNOME/Linux", or "Freedesktop/Linux". No one there talks about "GNU/Linux". You might love the BSD licencense more than the GPL but you know that without GNU, without choice of various licencense we all wouldn't be writing here today.

Red Hat is all about lock-ins. Don't believe me (maybe becaues Red Hat did really cool Linux advertisements a decade ago?), - they are *not* the good guys: http://www.forbes.com/sites/benkepes...g-open-source/

Next huge problem: Who is Red Hat's biggest customer? The US military ? this includes the US intelligence sevices. The problem with that is: they have an unbelieveable high amount of money. Their influence is just huge. The problem for me is *not* that it's the military of the US. (And I actually hate anti-American thinking)

The problem is their sheer influence. Not on Red Hat, but their influence through Red hat ON YOU. The US military complex and the NSA loves to recruit persons which have problems with their ego like Poettering. And you have to eat the shit Poettering serves you. He can say his "No one forces you to use it" bullshit as often as he wants. He can bitch about "You get it for frrrreee! You get it completely for frree" (I love how his German accent comes trough when he is angry. ^^ https://www.youtube.com/watch?v=_ERAXJj142o&t=3225) all the time. It's only lies. You ARE forced to use systemd. (except Slackware and Gentoo, and literally one or two others).

What Red Hat is doing is a Blitzkrieg against the open source community. Because it is not possible to implement another init system. It is just a lie with which they try to make you look folish or lazy. Systemd is shoved down everyone's throat. And we might have to ask, how the US military and the NSA and others have people in EVERY distribution. In EVERY telecommunications company, and so on. The surely never said "Oh, we just ignore Debian, or Ubuntu." The have their implants everywhere.

If you want a good read on that: http://igurublog.wordpress.com/2014/...cts-your-life/ (tinfoil-hat free)

Lennart Poettering is nothing less than the assasin of Linux distributions as we know them.

You might think I'm a douchebag? Let's talk again in 5-10 years. Major mainstream user-friendly distributions right now (with recent software packages) which barely use systemd functionality like e.g. Ubuntu 14.04 LTS and Debian 7.x., and Debian 6 LTS. Their support ends in a few years time. Ubuntu 14.04 in 2019. 16.04 will be conquered by systemd. Let's talk again in 2019.

[nslay]

Yes, I have been using single-user mode on UNIX machines before there was Linux. How about you?

I could care less about your 25 years of experience with UNIX. I'll translate some of your responses for you (obviously exaggerated, sorry):

I'm more experienced, so therefore I'm right.

Or

You've not been using it long enough, so you're wrong.

It's downright silly and irrelevant in this discussion.

Besides, your original post about the watered down reasons for the division between /bin, /sbin and /usr/bin, /usr/sbin gave me a different impression about your experience. For someone who is relatively a newbie like me compared to your 25 years of experience, that division is pretty obvious and reasonable. That said, based on your later responses, my initial impression was clearly unfounded (regardless of your posts about your purported experience).

[Paul Frederick]

To the actual hardware, a single executable binary for any particular init task is a hell of a lot simpler than running a larger general purpose interpreter binary with an ascii script as the argument it is called with. Simple to the user and simple to the hardware are often directly opposite oneanother. A happy medium would be for each module to have simple, easy to understand source that compiles easily to the final binaries without a lot of otherwise unused build dependencies.

Me being the user I have to take my own side on this issue. The hardware does not seem to be having a difficult time of handling things on its end either. All of this being the case I fail to understand why we need compiled binaries in the process. That to me is just an extra layer of obfuscation that serves no useful purpose. You're also overlooking the fact that systemd is 10 times larger than the current init system in terms of lines of code. So I'm not seeing how something 10 times larger is, a hell of a lot simpler, as you put it. Quite the opposite seems to be the actual case to me.

But if you just keep on saying things enough times they'll come true eventually right?